Your Client's ChatGPT Chats Aren't Privileged: A Former Federal Enforcer Explains

Collin McKee

Welcome back to the Endeavors AI Podcast. I'm your host, Collin McKee, founder of Endeavors AI, where we talk about the intersection of artificial intelligence, law, business, and the new economy being built around all of it. My guest today did his career backwards on purpose. Most lawyers finish in government. Dennis started there, at the HHS Office of Inspector General, working fraud enforcement in federal healthcare programs. Then Big Law. Then a national healthcare boutique. And last year, he opened his own firm, DSP Advocates, in New York. He runs it as a firm of one, on flat fees, with AI doing a lot of the work that a first-year associate used to do. And he's got a warning for every listener today about what happens when your client starts to use AI as well. Dennis Sapien-Pangindian, welcome to the show, sir.

Dennis Sapien-Pangindian

Thanks so much, Collin. It's a pleasure to be here.

Collin McKee

Awesome. So let's start where you started. Most people go into private practice first and land in government later. You went straight into the Inspector General's office. Why did you choose that path?

Dennis Sapien-Pangindian

Sure. I went into law school knowing I wanted to do healthcare law, but not really understanding what that meant, except for potentially medical malpractice. And I learned pretty quickly that isn't what I wanted to do. No offense to any med mal attorneys out there. But during my second year of law school, I found the whole wonderful world of healthcare fraud and abuse. It's the whole body of law and work associated with regulations on how healthcare and life sciences organizations should act and function, especially in the Medicare and Medicaid ecosystem.

I had an opportunity as a law student to intern in the government in Washington, D.C. I was at the Office of Counsel to the Inspector General for the Department of Health and Human Services, and that agency handles all matters of fraud, waste, and abuse for federal healthcare programs. So I thought, what better place to start than that? Luckily, I got an offer there. I knew I was going to be thrown into the fire if I went right into government. And again, no offense to anyone who goes the traditional path, but I knew that as a first-, second-, or third-year attorney straight out of law school, if I went into a big law firm, I would probably be doing what some would call scut work, right? I wanted to avoid that grind as well. So it was a win-win. I got thrown into the deep end of the healthcare regulatory enforcement world and avoided the grunt work of a more traditional legal path.

Collin McKee

Awesome. So can you tell us a little more about what you were working on day to day with the DOJ? You say fraud cases, but can you be any more specific?

Dennis Sapien-Pangindian

Sure. My bread and butter was False Claims Act cases. It's known as Lincoln's Law, because that statute goes all the way back to the Civil War, when the Union was tired of getting scammed by people selling weapons that didn't work, or horses that couldn't be ridden, to the Union Army. So they made it illegal to submit a claim for reimbursement from the federal government for an item or service that was fraudulent or defective. And you can imagine, however many years later, with Medicare and Medicaid being the largest single payer insurance provider in the United States, how relevant that can be, especially with how expensive healthcare is in the United States. It really is the hammer that DOJ uses to enforce the nation's fraud and abuse laws in the healthcare and life sciences ecosystem. And this isn't even a partisan issue. Eliminating fraud and abuse from federal healthcare programs is applaudable, and it remains a top priority for the Department of Justice no matter who's president. So I worked with DOJ on a lot of the civil False Claims Act cases and represented the agency in investigations, in settlement negotiations, and in exercising its own administrative authorities.

Collin McKee

Thanks for the explanation. So then you flipped it. In 2021 you went from building the cases to defending them, a complete one-eighty. What was the hardest adjustment in going from one side of the bench to the other?

Dennis Sapien-Pangindian

It was really interesting. When you're in the government, especially if you're being trained up as a fresh attorney, you're seeing all these cases where the government has a strong case. You're seeing the worst of the worst, and it gets very difficult to see what possible gray area there could be with these fraud and abuse laws. It seems clear-cut, binary: don't defraud Medicare and Medicaid, right? But you quickly learn there is a lot of gray area. A lot of these laws can be somewhat antiquated when you try to apply them in a current and dynamic environment, especially in a healthcare and life sciences ecosystem where there are a lot of new entrants trying to disrupt the marketplace and make it cheaper, more affordable, and achieve better patient outcomes, especially through the use of technology. So it was a steep learning curve. There's a lot more story to tell when you're living the day to day of an operational healthcare and life sciences client. It wasn't a surprise. I knew it couldn't all be bad. But actually being in it was definitely a learning curve.

Collin McKee

Right. So you're in your second step, working for a big firm, probably making good money, doing good work. What was the moment it popped into your mind that you had to go out on your own?

Dennis Sapien-Pangindian

I was at a big law firm, Paul Hastings, a global firm, and then I went to a national healthcare and life sciences boutique, Epstein Becker Green. It's nothing against those firms. I was on the partner track. I was senior counsel at Epstein Becker Green, and I was at the point where I was trying to build my own book of business, reaching out to clients and potential clients. What I found, especially in the legal world, was that the clients I really wanted to work with, the innovators, the disruptors, the new entrants in the marketplace, really couldn't afford traditional law firm rates. Especially by the billable hour, working with big teams and the really great partners at the firm, whose rates are so high. For a fresh-off-the-ground startup working out of a garage, that's difficult to swallow. But they still need to navigate this highly regulated space. So I thought, what an opportunity for me to use my knowledge and experience from government, and from learning from some of the best Big Law attorneys, to bring that value to people who are really trying to do good in this space.

Collin McKee

You bring up a very valid point. It's a pretty big problem, especially for those who are cost-conscious in their early days. That brings us to your fee structure, which I'm seeing more and more. You went from the billable hour to flat fees. You told me a little about why, but can you dig in further? Because who wouldn't want to get five hundred dollars an hour, or whatever the rate is? What's your why, and what made the business make sense in a flat-fee environment?

Dennis Sapien-Pangindian

Right. I wish I could say it was completely altruistic, Collin. But starting out, I did try the billable hour in the first few months of hanging my own shingle. And I realized, wow, this model kind of sucks, because I do all this work and I still have to wait thirty days before I can try to collect. I'm just starting this firm, I have no other salary coming in, and I'm burning through cash. So I reached out to some mentors of mine. One of them was a former professor from my law school. He's the one who put me onto flat fees. He said, "Look, no one's going to care that you charge fifty dollars less an hour. What they might be more enticed by is the value of a flat fee schedule. They don't have to worry, 'Do I call Dennis? The meter's running if I have a question.'" And oftentimes you leave money on the table with an hourly rate. A flat fee can be financially beneficial, but it also provides predictability and transparency to the client, so it's a win-win.

The more I experimented with it, the more I found that, one, it sets me apart as a differentiator in the market, and two, clients really love it, especially during consultations. They're already stressed. No one wants to have to hire an attorney, and they're calling me because they're stressed about a problem. I don't want to give them another problem to worry about with an unpredictable fee. So if I can scope out their project, I can quote them and say, "Look, this is going to be your flat fee, no matter what. I'll see you through this to the end, and you won't have to worry about me stopping work because you can't pay anymore." So as much as possible, I scope things out and do a flat fee, and clients have really loved it. And the use of technology and AI has really helped me scale my efforts to make that flat fee more profitable for me as well.

Collin McKee

Right. I've heard that before. It gives the client a sense of trust. They know what they're getting into. Communication lines stay open. To your point, they're not worried about how much a phone call is going to run them, and you can communicate. You touched on the AI portion, and I want to let you run with that. In our initial call you said flat fees are where AI really gets to earn its keep. I love how you worded that. Tell us how that works in your practice.

Dennis Sapien-Pangindian

Really, I use, and I'm not a paid sponsor or anything, I'm just a very happy customer, Clio for my practice management software. Last year, and it was perfect timing, they merged with vLex, the company behind Vincent AI, and got to incorporate a lot of the AI they acquired through that. So I have almost a one-stop shop in Clio. I have a CRM system to track my leads. I have my practice management system to track all my active, closed, and pending matters. And I also have a research system, so I don't have to pay the fees for Westlaw or LexisNexis, which can be killer. There's also Clio Work, which provides their research assistant for legal research. That has really been my secret weapon as a solo practitioner in making this all work, because I can quote things out.

It's kind of crazy to say "the old-fashioned way" now, without AI, but if I were researching things on my own, cold, it would take me almost a full day to do a proper, thorough search, to where I'm comfortable that I've exhausted my search and can draft the full memo or advise the client. Now I can get the same work, with the same confidence level, in one or two hours max. And with all the caveats: this is why I love Clio. It's a self-contained system, so it's not like ChatGPT, where I have to worry about it hallucinating a new case. It's only pulling from cases in its database that I know are real. So it's been a really great differentiator and a powerful tool for making the flat-fee system work. Of course, it has its caveats. I take a trust-but-verify approach with my AI tools. I pretty much treat it like a first-year associate, or maybe even a summer associate who's still in law school, that I don't completely trust yet.

Collin McKee

Yep. And we like to get the good, the bad, and the ugly, and be honest about it. In our first discussion you told me that when you started out on your own, hanging your own shingle, you were doing all kinds of things with AI, like bots for outreach, lead scraping, agents for contact info, and so on. So what happened? Tell us some of the low points, what you learned, what you did the wrong way and had to correct. Some of the horror stories.

Dennis Sapien-Pangindian

When I first tapped into AI as a new entrepreneur, I was listening to all these podcasts and everyone talking about how they're using AI in their practice. Like, oh, this is great, I could do it for this, I could do it for that. I played around with whether to just have a phone number that calls my office phone, or use an AI agent to answer my calls. I did a little trial run and realized pretty quickly that I don't want to do that. Same thing with lead scraping tools. I used AI to program a lead scraping tool that would do outreach, and it was kind of a bust. It would do the outreach, and there were almost no responses.

The more I thought about why these things weren't working, it came back to this: at the end of the day, I am the firm. It's not my tools. My tools are meant to support me. If I think about the ideal client journey: if I were stressed out and wanting an attorney, I wouldn't want to have to explain it all to an AI bot before I can talk to a human. No. I'd want the comfort of a person who's holding themselves out as a trusted guide, a sherpa through my tough times, not a cold bot. That made sense to me. I think of myself as a people person, and I enjoy that part of the work. So why would I want to cut myself out of that? That's one of the most foundational contact points where you can build a long-standing relationship with a client. It's that first contact.

So understanding where AI makes sense and where the human touch is more critical was really important. I use it less for my business development efforts, my outreach, my interactions with clients, and more for the internal work that needs to get done: legal research, maybe first drafts of work product, verifying citations. Even simple formatting has been great, because there's that old law student meme where you hit the space bar in a Word document and the formatting goes crazy, the bullets all change, the tabs are all indented, sirens in the distance. That's how it feels sometimes. AI really helps with formatting and takes away that little busy work.

Collin McKee

Yep, a hundred percent. That's what it's meant for. You said another thing that really stuck with me: that you do white glove service, not a metal hand. Why does that matter so much, especially for a firm your size?

Dennis Sapien-Pangindian

I think it's also about what I want my firm to be. It's DSP Advocates, not the Law Office of Dennis Sapien-Pangindian. It's not that at all. I want to be your advocate. I want to be your guy. I want to be Tom Hagen in The Godfather, your consigliere, right? And I don't want to be just another bot. You have all the tools already. If you really wanted to, you could use any of the cheaper services that are online now, LegalShield or anything else that'll draft a document for you in five seconds for ten dollars. And if I really wanted to be cold and cutthroat, I could try to do that and just have AI automate my firm. But that's not who I want to be. I want to be a guide for you. I want to be an advocate. Again, it goes back to why I went to law school in the first place. I want to help people who want to help people, and those people don't want another answering machine. They don't want a cold voice or an AI bot speaking to them. They want to know that a human being is hearing them, hearing their pain points, and giving them the one comfort that you're going to get through this. So it goes back to my philosophy of being a lawyer, an advocate. It's really being that for my client, and not just a cold tool they keep in the shed.

Collin McKee

Right. To that point, most people are exploring AI. I would hope everyone is at least wading into the pool. But a lot of your clients are using it as well, and that can be a problematic situation. What do you think is the first risk most people don't see in using AI in your field? If someone comes in and says, "Hey, I've thrown this into ChatGPT," what are they not considering?

Dennis Sapien-Pangindian

I think there are several things, but first and foremost: whatever they put into ChatGPT is not privileged. Their conversation with me is privileged. The contracts I have with my vendors, and the inputs I put into my AI, are privileged and protected as attorney work product, and they're secure. But if they're just putting something into ChatGPT, that's discoverable. There's case law coming up all across the country now that's establishing this, and that's because ChatGPT, Claude, all these AI tools, none of them are actual attorneys with whom you have an attorney-client relationship.

Especially if you're going to be getting into potential litigation, you don't want to be putting in all different queries and search terms that might indicate your state of mind, or expose you to potential liability just because you didn't word it correctly. That's all going to be discoverable, and it can show up in court. It's the same with email. I tell my corporate clients to have good digital hygiene: send every email as if a judge and a jury might see it one day. It's the same with your search terms. A lot of people come to me and say, "Hey, I'm putting this into ChatGPT," or they come with a really long, well-organized, thought-out Word document with the blue highlights, which is a key giveaway that it came from ChatGPT, and the bullet points, and sometimes even emojis. And I ask, "Did you get this from ChatGPT?" They say, "Yeah." And I say, "Well, just so you know, that isn't privileged, and I advise you not to do that moving forward."

The second issue is less about privilege and discoverability and more of a practical issue about the attorney-client relationship. If a client is constantly going back to ChatGPT to check my work, and then coming back and saying, "Oh, but have you thought about this? What about that?" I say, "Look, I'm giving you the expertise, experience, and knowledge. You're hiring me for that purpose. It's okay to question. Of course, question your attorney, ask good questions. But don't throw everything at the wall just because ChatGPT says you should do this or that, when I'm telling you it isn't a relevant, strong, or even legitimate legal theory." It can get frustrating. It can cause tension in the attorney-client relationship, and it can honestly confuse the client, who's getting information overload and doesn't know how to parse through it. I tell them, look, if you're going to hire me to be your guide, your advocate, your sherpa, then let me do that for you. I'm asking for a little bit of trust. So it's those two things, but definitely the imminent one is discoverability and privilege.

Collin McKee

Right. We're seeing that more and more across the board from all the attorneys I speak with. It does make it a challenge, and you've got to handle that conversation delicately, but professionally and firmly. Like, look, you reached out to me for a reason, and it wasn't to teach me how to do my job or teach me the law.

Dennis Sapien-Pangindian

Exactly. So use it as a way to inform. It's like a patient going to a doctor with a printout of every possible kind of cancer they might have because they have a cough or a tickle in their throat. That's the analogy I give my clients when they're getting a little overboard with AI.

Collin McKee

Yep, a hundred percent. Well, we're about at time. Let's wrap up with a couple of quick questions I like to ask everyone. What's one thing every solo or small-firm lawyer should start doing this week with AI?

Dennis Sapien-Pangindian

Do an AI audit. Check your terms and conditions. Make sure your accounts are enterprise accounts, where they're not being trained on your inputs. And do a few tests on it yourself before you start using it for your client matters. Make sure it's giving you outputs that actually make sense to you, and not hallucinating cases, things like that.

Collin McKee

That's great advice. I love that. So where do you think AI is headed in the next two years? It's hard to know where it's going in two months. But in your experience, where do you think we'll see AI in legal practice, say in a solo law firm, two years from now?

Dennis Sapien-Pangindian

Oh, we're going to see more and more people using AI. It's going to be the standard. I think attorneys are going to start getting trained on AI in law school. And honestly, I feel for law students now, because they're going to be competing with AI for jobs. Big firms are going to start hiring less. Companies are going to start hiring less. And, tinfoil hat on, it really worries me about the next generation of lawyers, who may not get trained up properly because legacy attorneys are relying more and more on AI as opposed to training new associates.

Collin McKee

It's a great point. I've heard it from so many guests who say they treat AI as a first-year, and you said it today too. So if you're a first-year law student with AI skill sets, that's a completely different thing, and that's value you can bring into a firm. That's really good advice. And lastly, where can people find out more about you and DSP Advocates?

Dennis Sapien-Pangindian

Sure. On my website, dspadvocates.com. I offer a free thirty-minute legal consultation that you can book with me via Zoom through the website. You can learn all about my practice, my background, the different types of work I do, my billing schedules, things like that. I also have a legal blog on there called Asked and Answered, where I try to answer typical questions clients may have that might be relevant to them. So, DSPAdvocates.com.

Collin McKee

Awesome. Creating value, I love that. That's great, Dennis. I appreciate the conversation. This has been great. Everything about DSP Advocates will be in the show notes; we'll link everything there for you. And if you're a firm owner wondering where your business actually stands with AI, if you feel like you're getting left behind, or if you just have questions about audits or SOPs, that's what we do here at Endeavors AI, so please reach out. My links will be in the show notes as well. For both Dennis and me, we really appreciate you listening. I'm Collin McKee, this is the Endeavors AI Podcast, and we will see you on the next one.